This project focuses on the study of membranes, proteins and carbohydrates by molecular dynamics computer simulation. Progress is reported under each Aim listed above Aim 1. Understand Model Membranes. This Aim involved extending our preliminary study of bending constants reported last year (Levine et al, JACS, 2014) to 12 different homogenous lipid bilayers using the CHARMM36 lipid force field. Excellent agreement with experiment was obtained for surface areas and area compressibility, and the trend that simulated bending constants agree nearly quantitatively with flicker experiments but not x-ray or pipette aspiration continues to hold. The evaluation of bilayer bending constants allowed extraction of the spontaneous curvature (from the first moment of the pressure tensor), and a comparison with these two quantities with those obtained from experiment on the inverse hexagonal phase. The spontaneous curvatures of the leaflets in bilayers are approximately 30% small than those in the inverse hex phase. This result is significant because values for the inverse phase are commonly used to interpret cellular events such as membrane fusion, and models should be reanalyzed with the more accurate estimate of the bilayer spontaneous curvature. (Venable, Pastor, Brown, Chemistry and Physics of Lipids, 2015) Aim 2. Develop Simulation Methodology. Progress has been made in: extending the CHARMM 36 force field to include calcium, cardiolipin, and PIP2; determining the importance of using P21 boundary conditions on peptide insertion into bilayers; assorted topics in diffusion, including effects of periodic boundary conditions in bilayer simulations, anisotropic diffusion in membranes, and effects of ensembles. However, there are no FY15 publications for this Aim. Aim 3. Simulate Complex Membranes Three papers related to this Aim were published, each covering a different topic. The curvature induction by the antimicrobial peptides (AMPs) piscidin 1 and piscidin 3 was determined in 4 different bilayer mixtures: 3:1 DMPC/DMPG, 3:1 POPC/POPG, 1:1 POPE/POPG, and 4:1 POPC/cholesterol. Importantly, the peptides induced positive curvature in the latter 3 systems, while negative curvature was induced in DMPC/DMPC. This result highlights the importance of membrane composition, and begins to provide clues into the action of AMPs (Perrin et al, J. Mem. Biol., 2015) The extension of the CHARMM 36 force field to palmitoyl sphingomyelin (PSM) reported in last year (Venable et al, Biophysical Journal, 2014) allowed us simulate liquid disordered and liquid ordered phases of PSM/DOPC/chol and PSM/POPC/chol, and compare the results with DPPC/DOPC/chol (also reported last year: Sodt et al, JACS 2014). The hydrogen bonding abilities of PSM lead to substantial differences with phases with DPPC, and it is likely that rafts in cells make use of the unique nature of each of these lipids. In particular, hydrogen bonding between cholesterol and the amide of PSM rotates the amide plane, which primes it for more robust bonding with other PSM. Cholesterol-PSM hydrogen bonding also locally modifies the hexagonal packing of hydrocarbon chains in the liquid ordered phase of PSM mixtures. (Sodt, Pastor, Lyman, Biophysical Journal, 2015; cover article for Sept 1 issue, and New and Notable) A simulation study of the mechanism of fencing of PIP2 by indicates that a functioning fence assembled from filaments of actin is unlikely. Fencing by septin is possible. However the filaments must be buried well below the membrane surface, have more than a single row, or contain additional components that fill small gaps in the filaments. (Lee, Im and Pastor, BMC Biophysics, 2014).